"The archive is corrupted" every time I do a scheduled backup

IF you are not backing up across thenetwork but on an individual machine, the most liekly cause is a spot of bad memory -- since ATI will tend to use up all the memory, a bad spot can show up that doesn't show with most programs. -- running memtest overnight is the best way to check this.

Other causes, if you are using a bootCD is a driver issue with the drivers onthe BootCD, which usese linux andlinux drivers rather than the windows drivers. INthis case, downloading a diff iso for a bootCD can sometimes get you a better driver. You sign into the Acronis web site and go to you registered programs page and you'l find a downloadable iso.

If you're trying to backup across the network, that's a bit out of my field and I must defer to others.

BTW, is the eror code showing up in the log? Anything else in the log? Is it showing up only upon attempted validation? If so, can you copy the backup to another machine and see if it validates ok?

The memory is fine. I have run and rerun Memtest86+ on it multiple times.

I will try validating on another machine.

Since posting this I have found a lot of threads on various forums on this topic. It appears to be an issue with ATIH2010 and no one has really figured out the true cause of it yet. I will dig more and see what I can identify. I'm usually pretty good at isolating problems like this.

Scott

If saving over the network, one test that may be helpful in trying to figure out where the problem is would be copy a large file (several GB's, if possible) from the computer to the network share and then see if the files are identical (binary file compare or MD5 checksum, for example).

Is the validation failing immediately or does it run for a while first? Immediate errors are usually caused by the file not being read/detected correctly and not the file itself. If it fails after running for a while, does it always fail in the same place or does it vary (sometimes 10%, sometimes 35%, etc.)?

It runs the validation for a while before it fails. I am not sitting down an watching it so I don't know exactly when it happens.

As a side note. I like your large file network test using MD5 checksums. I have never thought of that. Good tip!

I've used that method before to pin down an intermittent (at first) RAM error on my WHS. Large files wouldn't compare correctly. Later, even small files (100MB) wouldn't match. If it passes okay after several times (5-10), it's probably not the network or the RAM (as used by the transfer).

If you have the TI logs, you could look at them and check the times for the start of the validation and the failure. This would give you a good idea if it's the same or not. If the time is moving around, it can mean you're getting network errors, read errors, RAM errors, etc. It can also just mean that the network connection speed isn't consistent (this happens to me a lot).

Also, some netowrks throttle transfer rates if a user starts trying tomove lots of data. However, since the prob would then show up on all the machines.

I did a lot of testing this afternoon and early evening. I have isolated it to the machine that I primarily use. I am suspecting a SATA problem, probably cable, but for the time being I'm not 100% what the problem is.

I was pretty heavy handed in my tests on all machines on my network. I did 25GB file transfer to and from my NAS (Synology DS1010+) and on each transfer I did an MD5 checksum. Each transfer took about 8min. I did this at least 4 times per file (I used three files per machine). I also did a lengthy series of MD5 checksums on very large files on the hard drive and did not copy of move them on each machine. Finally, on the machines that passed these MD5 checksum tests both over the LAN and on the local disks I did several full disk backups and validations with ATIH2010. All but one of my machines passed all of these tests. So it wasn't ATIH2010 causing the troubles, it was my one machine as several of you mentioned. However, it was not related to doing a backup over the network.

As for the machine having the problems. I cannot even get consistent MD5 checksums on the same file of very large size. I have swapped SATA cables and motherboard SATA ports several times and each cable/port has slightly different rate of results. I have also run Memtest86+ v4.00 through two default passes without error. When I nail this down, I will post what I have found.

Scott

Is the MD5 failing to match whether the file is on the internal drive, NAS, USB drive, etc. or is it just from the internal drive?

Even though Memtest is not showing an error, you might do some RAM swapping and see if anything changes. You might swap the sticks (if dual-channel) or rotate them (if triple-channel). If you can, take some out and run a few tests (drop from 4GB to 2GB, for example).

You could also check timings and the voltage. I've had a computer change these without my knowing. Some of these problems can be really hard to find.

With this machine it's failing regardless where the file is located. I'm guessing that's why you're thinking it could be the RAM, huh?

That's right. I don't know the computer setup, but that's what I would play with first. You could relax the timings and see if that makes a difference. That's easy to do in the BIOS without needing to make any hardware changes.

Well.. normally you would think it would be easy. Sorta long story, but this is a Dell Inspiron 530S and they have blocked just about everything from the BIOS including being able to enable virtualization on the CPU. I have built over 15 computers and bought prebuilts twice. After the first prebuilt I bought I vowed to never buy prebuilt again because of the hell I went through to get it working right. I have since built roughly 15 home rolled machines that have worked great. Then my wife needed a new machine recently and I thought I'd just get something simple that was ready to go right out of the box. HUGE mistake. Ok.. so on with what I know.

I upgraded the RAM a while ago using http://www.newegg.com/Product/Product.aspx?Item=N82E16820231098. The timing of the RAM in the Dell is 5-5-5-12 and this G.Skill is 5-5-5-15 so you'd think that is the problem and a boneheaded mistake on my part, and maybe it was. According to what I read the voltage tolerances on the Dell are very tight and specific and that's why most people used this RAM when upgrading and said it had no problem meeting the 5-5-5-12 timings. So now I'm thinking that's BS. I have been pretty nerdly about computer building in my day and have never encountered a RAM problem that Memtest86+ couldn't find. So this is all new to me. I'm now starting to have some theories why a continuous series of MD5sums on the same file for several hours catches other RAM errors that Memtest86 doesn't.

It's been running all night on the same 35GB file and it is 100% success rate on the MD5 sums. I'm confident this solved my problem, although based on the other results I might also have some flakey SATA cables to deal with, but I'll isolate that later now that I have a working baseline.

Thanks for helping with this. When I wanted to use XP Mode on this machine before MS had the patch so it could workout without HW support for virtualization I learned of a BIOS hack that someone did for this particular machine that enabled all of the typical BIOS settings. It is based on a slightly older version of the Dell BIOS but it's probably new enough for my needs. Since it seemed like only 3 people comments on using it I was a little nervous about attempting to flash this machine with some unknown BIOS so I just dealth with it until MS released the Virtual PC patch. Now I have another reason to consider that jacked BIOS. Just not sure if it's worth the risk.

Thanks for the help. I tell ya... I thought I knew just about everything about PC hardware because if there's a problem that someone has had, I've had it already. This is a new one for me and now I have another tool for my bag of tricks to bang on a system to see what shakes. Thanks for that!

Cheers,
Scott

PS - I know this is more than you care to know. I just want to set the record straight that I don't normally buy prebuilts.

... It's been running all night on the same 35GB file and it is 100% success rate on the MD5 sums. I'm confident this solved my problem, ....

I'm not clear what solved the problem.

My experience says that marginal RAM errors can be more frequently detected by the random data tests than the walking bitmap pattern tests. If I don't get a failure after a couple of passes I will run the random data test only. While this used to be a very well known effect with magnetic core memory I think that data dependent errors exist in semiconductor memory systems. The issue is that there can be a greater or lesser probablility of failure depending on what the previous data value was in either the chip or the stick or the cache. So if a cell is just on the edge some data pattern sequences may cause a failure where all others won't.

I won't buy prebuilts either unless its a notbook of course. Throttled BIOS for the masses, crapware loaded, recovery rather than real OS DVD, etc.

I think it came down to heavy traffic on the bus to the RAM that led to timing errors, a scenario that doesn't happen when using walking bit patterns tests like Memtest86+.

Here's an example of the results I got on this large file doing the repeated MD5sums just so you can see it took a while in certain situations to see a problem.
7dff3c0618947b7bfb376569992e7aa8
7dff3c0618947b7bfb376569992e7aa8
7dff3c0618947b7bfb376569992e7aa8
a110d4892a0e2de4cc6f8fec373867d0
7dff3c0618947b7bfb376569992e7aa8
5339262e40de5a196cde1b83068e8c83
7dff3c0618947b7bfb376569992e7aa8
7dff3c0618947b7bfb376569992e7aa8
7dff3c0618947b7bfb376569992e7aa8
8a29dc6f84f16aad5bd26db6e1441f2e

I have tried validating the system partition using the rescue disc, and it fails every time.
I tried creating the image both on the external USB drive as well as directly to the HDD.

This issue affects such a small percentage of Acronis True Image Home 2010 users, so I doubt whether they'll be a fix until True Image Home 2011. I have invested a huge amount of time into getting software to work. I've come to the end now. I have lost all confidence with the product. I just can't trust it any more as a professional backup product.

But there is light at the end of the tunnel...

I stumbled upon Paragon Backup & Recovery FREE Edition:
http://www.paragon-software.com/home/db-express/download.html
Does the same job as Acronis True Image Home 2010 without the validation hiccups.
And it's FREE. Been using Paragon software for years on the server and it's excellent.

Good luck the rest of you..
Over and out.

Hello all,

Scott, MudCrab and Seekforever, thanks for your help, we really appreciate it.

Mark, I may suggest you to create a backup of your system partition and try to validate it using a bootable media. Please let us know the results of the validation.

We are looking forward to hearing back from you at your earliest convenience.

Thank you.

Solution to Event code: 0x000101F6 Operation with partition was terminated

Details: Operation with partition '0-0' was terminated.
Details:
The archive is corrupted. (0x70020)
Tag = 0xF5F8CBCF76155639

More information can be found at: http://kb.acronis.com/errorcode/
Event code: 0x000101F6

I hope this can get published somewhere to help everyone out there ... including Acronis & maybe even me Hint hint (free software or something perhaps) hint hint ... as so far, in all my searching I haven't come across the following solution.

I must point out however that for "try and decide" in TI Home 2010 Acronis does warn to turn off background defragmentation.

I was having terrible times getting any kind of reliability with validation, or mounting of images. Random failures galore including a restore that went terribly wrong and left a Netbook unbootable (Netbook = NO optical drive so had to create bootable USB Key etc etc I have much less hair now !!!)

Anyhow ... I was backing up using True Image Home 2010 on Windows 7 starter edition 2 GB ram to a Lacie 1 TB USB external Hard Drive. Never even thought of it but had the Lacie supplied "USB Boost" program running in boost mode. (does it automatically when drive is present)

Since turning that off while using True Image No more problems

I will wager that all these "USB Boost" type apps cause similar problems.

Hope this gets looked into and solved. But after many frustrating hours of troubleshooting I now have 100% reliability

Kami,
Vancouver
Canada